In this biography
Quick facts
- Born
- 1923
- Died
- 18 April 2003
- Birthplace
- Portland, Dorset, United Kingdom
- Country
- United States
- Nationality
- Not independently established
- Known for
- Relational model
- Early computing system
- STRETCH
- Collaborator
- Chris Date
Why this life matters
He separated a logical account of data from particular storage arrangements.
Timeline
Key moments
Computing career
Joined IBM.
Research record
Published the relational-model paper.
Recognition
Received the Turing Award.
From Britain to North American computing
Codd studied mathematics and chemistry at Oxford and served in the Royal Air Force during the war. He moved to the United States in 1948 and joined IBM in 1949. A period in Ottawa preceded his return to IBM in 1957.
His research emerged from experience with large computing systems, rather than from a database product already waiting to be renamed.
Sources for this section: [2]
Systems work and a different data model
He worked on multiprogramming for STRETCH with colleagues including E. S. Lowry, J. E. McDonough and C. A. Scalzi, and completed doctoral study at Michigan.
The 1970 relational paper described data through relations and operations rather than physical navigation paths.
That shift asks designers to specify what information a record represents before choosing how to traverse or store it.
Languages, dependencies and disagreement
His subsequent work included ALPHA, relational algebra and calculus, and data dependencies. He participated in debates over competing database approaches and later worked closely with Chris Date.
A logical proposal, a query language and a commercial implementation are related but distinct achievements. Neither an early paper nor a later product should erase the work between them.
Sources for this section: [3]
A lasting design question
He received the Turing Award in 1981.
He died on 18 April 2003.
The linked study follows a small but substantive consequence of the model: two binary projections may preserve individual pairings while losing which three values originally belonged together. A join can be valid as an operation without recovering the original relation.
Quick answers
Frequently asked questions about Edgar F. Codd
- Does every join reconstruct the relation that supplied its projections?
- No. The original example shows additional tuples appearing when a three-way association is lost.
Keep exploring
Lives connected to this story
- Barbara Liskov
Compare another approach to specifying software abstractions.
- Ken Thompson
Read a different systems-development career.
Sources and further reading
Explore the evidence
These references support the biography and offer further reading. Official records, first-person accounts and independent reporting provide different perspectives; image records document attribution and licensing.
- 1. American Academy of Arts and SciencesEdgar Frank Codd
- 2. IBM historyEdgar F. Codd: relational database pioneer
- 3. C. J. Date, NAE Memorial Tributes 12, 2008Edgar F. Codd
- 4. E. F. Codd, CACM 13(6), 377โ387, 1970; retained retypeset copyA Relational Model of Data for Large Shared Data Banks
The acquired paper is a retypeset academic copy. Selected operations and join text were read, not every proof or a historical facsimile.